VSP Parity Group Usage Test

A parity group is a group of hard disk drives (HDDs) that form the basic unit of storage for the Virtual Storage Platform. All HDDs in a parity group must have the same physical capacity. The parity group is also called as an array group or a RAID group. A RAID group contains both user data and parity information. This allows user data to be accessed in the event that one or more of the drives within the RAID group are not available. If one or more parity groups are inaccessible, then the load to the other parity groups may increase manifold! This sudden increase will affect the data transfer between the parity groups, increase the I/O traffic and result in irregular load on the parity groups which may sometimes lead to critical data loss thus affecting the overall performance of the storage device. To avoid such problems, administrators can use the VSP Parity Group Usage test.

This test monitors the I/O traffic and data transfers conducted by each parity group on the storage device, and indicates irregularities in load balancing across the parity groups.

Target of the test : A Hitachi VSP storage device

Agent deploying the test : A remote agent

Outputs of the test : One set of results for every parity group on the Hitachi USP device monitored

The figure below depicts the test configuration page of this test. Click on the parameters in the figure below to know what they are and how to configure them.

Measures made by the test:
Measurement Description Measurement Unit Interpretation

I/O operations rate:

Indicates the number of read-write operations performed on this parity group per second.

IOPS

A high value for this measure is indicative of high I/O activity on the parity group. Comparing the value of this measure across parity groups can accurately reveal which parity groups are overloaded, and also enable administrators to easily detect irregularities in load distribution across the parity groups.

To uniformly balance load across parity groups, you should consider installing additional HDDs, or you can use volume migration to migrate volumes from high-usage parity groups to low-usage parity groups.

Transactions rate:

Indicates the rate at which data transfers occur on this parity group.

KB/Sec

 

Read IOPS:

Indicates the rate at which read operations are performed on this parity group.

IOPS

 

Write IOPS:

Indicates the rate at which data is written to this parity group.

IOPS

 

Read hits:

Indicates the percentage of read requests serviced by this parity group.

Percent

Ideally, the value of this measure should be high. A low value is a cause for concern, as it indicates that a majority of read requests have failed.

Write hits:

Indicates the percentage of data written to this parity group.

Percent

 

Cache-to-disk transfers:

Indicates the number of data transfer operations performed from the cache to this parity group.

Number

A high value for this measure is a sign of good health. A low value or a consistently decreasing value could be a cause for concern, as it indicates that the cache is not writing enough data to the disk; this in turn could overload the cache and hamper its ability to make space for data that is waiting to be written.

Response time:

Indicates the current responsiveness of this parity group to I/O requests.

Microseconds

Ideally, the value of this measure should be low.

Transfers between disk and cache:

Indicates the rate at which data transfer operations are performed between this parity group and the cache.

Number/Sec